Clin. exp. Immunol. (1990) 80, 56-61
Quantitative measurement of cytomegalovirus-specific IgG and IgM antibodies in relation to cytomegalovirus antigenaemia and disease activity in kidney recipients with an active cytomegalovirus infection M. VAN DER GIESSEN, A. P. VAN DEN BERG, W. VAN DER BIJ, S. POSTMA, W. J. VAN SON* & T. H. THE Departments of Clinical Immunology and *Renal Transplantation, University Hospital Groningen, The Netherlands
(Acceptedfor publication 30 October 1989)
SUMMARY In a longitudinal investigation 103 kidney recipients were studied with respect to the development of cytomegalovirus (CMV) specific antibodies of the IgG and IgM class, in relation to the detection of CMV antigenaemia (immediate early antigen, IEA), in weekly obtained blood samples during the first 3 months after transplantation. In 15 out of 49 (31 %) seronegative patients a primary infection occurred, which was characterized by a quick rise in IgM antibody followed by a slower production of IgG antibody, high maximum numbers of IEA+ cells, and a CMV syndrome in 11 patients. In 35 out of 54 (65%) seropositive patients a secondary infection occurred. After a post-operative fall in the IgG antibody, which was also found in patients without an active infection and which was accompanied by a similar drop in serum albumin and IgG, a second dip in IgG antibody was found 6 days before the first IEA+ leucocyte appeared in the blood. This was followed by a significant increase, indicative of an active immune response in consequence of the infection, 18 days later. In 31 of these 35 patients an IgM response was found. This could be ascribed to the presence of rheumatoid factor activity in 20 of them. Eight patients who showed a transient rise in IgG antibody between the two dips could be distinguished from the remaining ones by a lower maximum number of IEA+ cells and less severe disease symptoms. The described results suggest that (i) an adequate humoral immune system may prevent symptomatic CMV disease in secondary infections; and (ii) CMV-specific antibodies may be removed from the circulation by antigens present in infected tissues before CMV antigenaemia becomes detectable.
Keywords cytomegalovirus antigenaemia test antibody
response
Diagnosis of an active CMV infection is hampered by the fact that clinical symptoms are non-specific and may not distinguish it from other infections or from rejection. The diagnosis of a CMV infection therefore is dependent on laboratory parameters, i.e. isolation of virus from clinical samples such as blood, urine and saliva, and the detection of a specific antibody response. Recently, a rapid and reliable method for the early diagnosis of an active CMV infection has been developed in our laboratory (van der Bij et al., 1988a, b). With the aid of monoclonal antibodies against the major 67-72 immediate early antigen (IEA) protein of CMV-infected cells produced in our laboratory, this antigen could be detected in peripheral blood leucocytes, even before clinical symptoms of the infection became manifest. A relation was found between the level of CMV antigenaemia and clinical disease activity (van der Bij et al., 1989).
INTRODUCTION Active cytomegalovirus (CMV) infection is a major cause of morbidity and mortality in renal transplant recipients (Glenn, 1981; Rubin & Colvin, 1986). This may result from reactivation of latent virus in the recipient due to the immunosuppressive treatment or from transmission of the virus from the graft if this is obtained from a seropositive donor, in primary as well as in secondary infection (Grundy et al., 1988). The incidence and severity of CMV disease is higher in seronegative recipients than in patients with a pre-transplant immunity against CMV (Pass, Griffiths & August, 1983; Smiley et al., 1985; Rubin & Colvin, 1986). Therefore, knowledge of the serostatus of the kidney recipient is important. Correspondence: Dr M. van der Giessen, Department of Clinical Immunology, University Hospital, Oostersingel 59, 9713 EZ Groningen, The Netherlands.
56
Cytomegalovirus-specific antibodies This finding and the above-mentioned influence of pretransplant immunity on the clinical manifestations of the infection prompted us to measure in detail the antibody response against CMV in these patients in relation to CMV antigenaemia and severity of the disease in order to gain insight into the relevance of the humoral immune response for the course of this virus infection. To this end an ELISA was developed for the quantitative assessment of CMV-specific antibodies, which was based on the method described by Middeldorp, Jongsma & The (1984). This method was applied for the sequential quantification of both IgG and IgM antibodies. MATERIALS AND METHODS Patients 'In the period between April 1986 and December 1987 103 patients receiving a cadaveric renal allograft could be longitudinally studied with respect to their humoral immune response against CMV. Serum was obtained on the day of transplantation and at weekly intervals thereafter for at least 3 months. After this period blood was obtained at irregular intervals. The initial immunosuppression regimen consisted of cyclosporin A, combined with low-dose prednisolone in case of a second transplant or in the presence of anti-HLA antibodies. In five patients this regimen was preceded by an induction course of monoclonal antibodies (OKT3) with azathioprine, in one patient initial immunosuppression consisted of anti-thymocyte globulin. Allograft rejection was defined as an increase in serum creatinine of > 20%, combined with one or more of the following: a swollen, painful kidney; fever; decreased urinary output; and salt retention or findings compatible with rejection in a renal biopsy. Anti-rejection therapy in patients on cyclosporin A-based regimens consisted of a bolus of 1000 mg methylprednisolone given intravenously on 3 consecutive days; in patients on an azathioprine-based immunosuppressive regimen rejection was treated with 200 mg oral prednisolone, in 3 weeks slowly tapered to the original prednisolone dose. Treatment with anti-thymocyte globulin was instituted after failure of first-line anti-rejection treatment. In case of documented vascular rejection six courses of plasmapheresis were given. CMV syndrome was defined as the occurrence of unexplained fever for at least 3 days in combination with a laboratory diagnosis of active CMV infection, i.e. seroconversion or a significant increase in IgG antibody and/or detection of CMV antigenaemia or viraemia, and one of the following: arthralgia; leucopenia (WBC < 3-10 exp 9/1), thrombocytopenia (platelets 50 U/i) or pneumonitis without other cause (Cheeseman et al., 1979). Detection of CMV antigenaemia (IEA test) This test was performed exactly as described by van der Bij et al. (1988a,b). Briefly, leucocytes were isolated from EDTA-blood, obtained at weekly intervals after transplantation for 3 months, by dextran sedimentation and cytocentrifuged (Shandon) on microscopic slides. The presence of CMV-IEA in the granulocytes was detected by immunoperoxidase staining, using a mixture of two monoclonal antibodies directed against the major 67-72 kD IEA antigen, horseradish peroxidase labelled rabbit anti-mouse immunoglobulin antibodies (Dakopatts) and
57
as substrate H202+AEC (amino-ethyl-carbazol). Results are expressed as the number of positive cells, i.e. cells showing nuclear staining, per 50 000 cells screened.
Preparation of CMV antigensfor serology Batches of CMV-LA (late antigens) were prepared exactly as described by Middeldorp et al. (1984) from fetal fibroblasts which were infected with AD169 (1 PFU/cell), harvested after 6-7 days, and extracted with an alkaline glycine buffer. A control antigen preparation was obtained from mock-infected fetal fibroblasts from the same batch. Determination of CMV antibodies (CMV-ELISA) Wells of microtitre plates (Greiner) were coated with 100 yl of a solution of CMV-LA antigens in 0-1 M carbonate buffer, pH 9-6. Optimal coating concentration of different batches was determined by block titrations, and always resulted in a coating concentration of 4-5 ,g protein/well. In each plate, wells coated with the control antigen preparation, used at the same protein concentration, were included to detect non-specific (IgM) binding in patients' sera. Plates were left at 4°C for at least 48 h. After washing the plates (Titertek microplate washer) with wash buffer (0 3 M NaCl, 0-01 M Tris, pH 8-0, 005% Tween 20), the wells were filled with 100 p1 of two-fold dilutions of the serum samples diluted in wash buffer + 1% BSA, starting with 1/100, and the plates were incubated at 37°C for 1 h. In each plate two rows were filled with dilutions of the standard serum. After washing the plates, the wells were filled with 100 p1 of the appropriate dilution of horseradish peroxidase conjugated goat anti-human IgG or IgM (Kallestad) and incubated at 370C for 45 min. After washing the plates the wells were filled with 100 MI substrate solution (20 mg OPD, orthophenylene diamine, in 100 ml phosphate buffer, pH 5-6, and 0-0045% H202 added just before use). After 15-20 min incubation at room temperature the reaction was stopped with 2 N H2SO4 and OD at 492 nm was read with a Titertek Multiscan. Results of patients' serum samples were calculated in relation to the standard serum using an on-line computer and a program developed in our department, based on log-logit transformation of the measurements (Rodbard & McClean, 1977). The results of positive IgM tests, which might be due to the presence of IgM-rheumatoid factors (RF), were always checked after absorption of the serum with a sufficient amount of pansorbin (Calbiochem) onto which heataggregated normal human IgG was coated. The pansorbin binding capacity was not saturated so that partial absorption of patients' IgG was still possible (resulting in about 50% reduction of the IgG antibody content), in view of the preferential binding of some RFs to patients' own IgG (Champsaur, FattalGerman & Arranhado, 1988; and unpublished observations). The quantity of pansorbin routinely used for absorption was calculated to be sufficient to remove all of the RF activity generally present. This was checked with a sensitive RF-ELISA and sera with known RF concentrations (kindly done by Miss Rina de Jong, Department of Immunochemistry). Standard sera Standards for CMV antibodies of the IgG and IgM class respectively were prepared from mixtures of different serum samples and put at 100 U.
M. van der Giessen et al.
58
For the IgG standard two volumes of serum from healthy controls with a high normal level of IgG antibody were mixed with one volume of a mixture of several patients' sera with high IgG antibody levels due to a secondary CMV infection. This mixture did not contain detectable IgM antibody. For the IgM standard, two volumes of serum from a normal seronegative individual who showed no aspecific IgM binding to either the CMV-LA or the control FF antigen preparation were mixed with one volume of a mixture of several serum samples from patients with a primary CMV infection which were obtained before the IgG antibody reached undesirably high levels. Thus prepared the IgM standard contained 4 U IgG antibody. Both standards were kept frozen at - 80'C in small aliquots.
Normal sera Normal sera were obtained from healthy blood donors (n = 84) and laboratory personnel and students (n = 27) to determine the local incidence ofCMV infection and the normal mean level and range of IgG and IgM antibody concentrations in seropositive individuals when using our quantitative ELISA assay.
Statistical analysis Differences between groups of patients and normals were statistically analysed using Student's 1-test, Wilcoxon's rank sum test, and x2 test. Correlation analysis was performed with the Spearman rank correlation test. RESULTS CMV-ELISA Reproducibility of the ELISA results was investigated by duplicate and triplicate measurements of 92 randomly chosen serum samples containing IgG and/or IgM antibody. Intraassay variability appeared to be negligible. Inter-assay variability was shown to be less than 15%. In view of this finding and the clinical data on active CMV infection of the patients an increase in IgG antibody concentration of 50% in relation to the previous value was considered to be conclusive for a secondary infection. CMV antibodies in healthy controls Results obtained with serum samples from healthy individuals, among them known seronegative and seropositive persons previously determined by immunofluorescence (The, Klein & Langenhuysen, 1974), showed that seronegatives always had less than 2 U IgG antibody in the quantitative ELISA. Therefore individuals with an IgG binding > 2 U were considered seropositive. With respect to the IgM-ELISA it was found that although most normals, seropositive as well as seronegative, had IgM levels < I U compared with the standard, some seronegative individuals showed weak aspecific IgM binding, which never exceeded 4 U, so an IgM antibody level of 5 U or more in a previously seronegative individual was considered to be decisive for a seroconversion. The prevalence of CMV IgG antibody in our series was found to be 53% (59/111). The mean level of IgG antibody concentrations, which were log-normally distributed, in these 59 seropositive individuals was 17 U with a range of 4-72 U.
Table 1. Incidence of cytomegalovirus infection in kidney transplant recipients
Patient
n
Infection
%
Seronegative Seropositive
49 54
15* 35t
65
31
* Immediate early antigen (lEA) test, n= 15, IgM antibody test, n= 14. t IEA test, n = 31, significant rise in IgG
antibodies, n = 35.
0) -0
.0
t2
9
IEA+cells
16 23 30 37 44 51 58 65 72 Days post-transplant 5 167 142 86 2 2 100 days after transplantation, and the exact date of their first positive IEA test could not be determined. The mean period between transplant and the first positive IEA test for the remaining 29 patients was 33 + 13 days. A significant increase in IgG antibody followed at day 45 + 10 after transplant.
59
Cytomegalovirus-specific antibodies Table 2. Concentration (%) of pre-transplant levels in 26 kidney recipients with and without cytomegalovirus (CMV) infection
5001- (a)
100
Time 8 days 4 weeks
n
Total protein
26 76+4
Albumin
26 72+5
IgG Non-infected CMV infected CMV antibodies (IgG) Non-infected CMVinfected
10 70+9 16 70+9
81 +6 80+9
91 +9
80+ 14
87+ 17
68+13
87+31
10 71+13 73 + 13 16 71+13 63+18 *
1 ,~~~~~~~
50
8 weeks
88+ 12
IC
)~~~~~~~~~~~~~~~~~~~~ 7
f
14 21
28 35 42 49 56
Days
78
post-transplant
IEA+ cells
